3/9/2026
Show HN: TubeTrim – A local YouTube summarizer using Qwen in pure Python
TubeTrim: Python tool summarizing YouTube videos locally using Qwen LLM without APIs or data leakage.
3/9/2026
Show HN: Ducktel – observability when the consumer is an AI agent, not a human
Observability tool designed for AI agents as consumers. Ingests OTLP, stores Parquet, exposes DuckDB SQL via CLI for agent reasoning without human dashboards.
3/9/2026
Show HN: Llmpm – NPM for LLMs
CLI package manager for open-source LLMs. Install, run, and share models like npm. Enables reproducible LLM project setup.
3/9/2026
Muon on Graph Neural Networks
Analysis of optimizer impact on neural network geometry in GNNs, examining weight matrix properties and information flow.
3/9/2026
How do you track and optimize your AI API spend?
Techniques for reducing LLM API costs: model routing (55%), prompt compression (70%), request deduplication (15%), semantic caching (20-30%).
3/9/2026
Show HN: Connect your research data easily to AI agents
Data layer platform enabling AI agents to access and analyze large unstructured scientific research datasets through novel ingestion algorithms.
3/9/2026
Show HN: A 2000s-style web forum where AI agents and humans hang out
Retro forum where AI agents and humans interact without goals or karma. Open API with Grok, Claude, Kimi agents.
3/9/2026
Show HN: Styx, Open-source AI gateway with intelligent auto-routing (MCP-native)
Open-source AI gateway with intelligent auto-routing across providers (OpenAI, Anthropic, Google, Mistral). MCP-native, cost-optimized model selection.
3/9/2026
Show HN: Locode, a local first CLI that routes tasks to local LLMs or Claude
Locode: open-source CLI routing simple tasks to local LLMs and complex reasoning to Claude for token efficiency.
3/9/2026
Show HN: Polpo – Build zero-human companies. Open source
Polpo open-source platform for building AI agent teams with roles and skills to autonomously complete tasks end-to-end without human intervention.
3/9/2026
Show HN: NVFP4 on Desktop Blackwell – 122B MoE on a Single RTX PRO 6000 31 tok/s
Running 122B Qwen 3.5 MoE model with FP4 quantization on single RTX PRO 6000 desktop GPU at 31 tokens/sec.
3/9/2026
Show HN: Bear – enforcing architectural boundaries for agent-generated code
BEAR: CLI governance tool enforcing architectural boundaries for AI agent-generated code in CI/CD pipelines.
3/9/2026
A job ad for Agentic AI Advocate
Job posting highlighting emergence of autonomous AI agents building and deploying applications independently.
3/9/2026
Large-Scale Online Deanonymization with LLMs
LLM agents successfully deanonymize users across HN, Reddit, LinkedIn using unstructured data at scale with high precision.
3/9/2026
Show HN: I indexed 45k+ AI agent skills into an open source marketplace
SkillsGate: open-source marketplace indexing 45k+ GitHub skills for Claude Code, Cursor, Windsurf agents. Semantic search with LLM-generated metadata.
3/9/2026
Replacing file systems with a canvas produced SOTA agent performance
Research showing canvas-based context replacement for file systems achieves state-of-the-art agent performance.
3/9/2026
Show HN: A-Life Powered by LLMs
Artificial life project using LLMs with cybernetic agents in file system environment, simulating evolution with neural network replacement.
3/9/2026
Show HN: Service marketplace for AI agents, 300 APIs and data sources
Marketplace connecting AI agents with 300+ APIs and data sources for task execution.
3/9/2026
LLMs Keep on Getting Better in Small Steps
Technical analysis of recent incremental improvements in LLM training: architectural details, optimization, and scaling.
3/9/2026
88% of companies use AI. Only 13% trained anyone how
Research shows 88% of companies use AI but only 13% trained staff. Highlights gap between AI adoption and practical knowledge.
3/9/2026
Jetbrains: Air Launches as Public Preview – A New Wave of Dev Tooling
JetBrains Air: agentic development environment in public preview. IDE designed for AI-assisted software development.
3/9/2026
Show HN: Wolf Defender, a open-weight prompt-injection detection model
Wolf Defender: lightweight open-weight BERT model for prompt injection detection and data leakage prevention.
3/9/2026
Show HN: DenchClaw – Local CRM on Top of OpenClaw
Self-hosted CRM built on OpenClaw agents for automating enterprise workflows like outbound calling and legal intake.
3/9/2026
Show HN: AriaType – Privacy-first voice keyboard with AI polish (Beta, macOS)
Local-first voice keyboard for macOS using Whisper and optional LLM polish, processing entirely on-device without cloud dependency.
3/9/2026
Show HN: A step debugger for AI agents
Step debugger tool for OpenClaw agents that simplifies tracing and understanding agent behavior through visual inspection of traces.
3/9/2026
Show HN: Skilo – Share agent skills with a link, no repo required
Tool for sharing AI agent skills via shareable links without repository requirements, supporting Claude and other AI platforms.
3/9/2026
Show HN: Screen-watching AI needs a kill switch
Desktop app using vision and MCP to give AI assistants context about user screen activity with local storage and privacy controls.
3/9/2026
Do AI coding agents improve velocity and quality?
Empirical exploration of whether AI coding agents improve developer velocity and code quality, with practical usage examples.
3/9/2026
The first AI agent worm is months away, if that
Analysis of emerging AI agent security risks, including malicious 'claw' style agents and potential AI worm threats.
3/9/2026
Let's be honest about AI Coding
Article examining effectiveness and limitations of AI coding agents for developer velocity and code quality.
3/9/2026
Show HN: Warden – access cloud APIs without giving workloads credentials
Identity-aware egress gateway replacing cloud credentials with zero-trust access for workloads and AI agents.
3/9/2026
Show HN: I had Claude rank every YC W26 startup
Claude used to rank Y Combinator W26 startups by scraping LinkedIn, press, and verifying product existence.
3/9/2026
Show HN: OpenClaw CRM, an open source CRM your AI agent can manage
Open source CRM designed for programmatic control by AI agents via OpenAPI, with skill integration.
3/9/2026
Show HN: Run autoresearch on a gaming PC (Windows and RTX GPUs fork)
Fork of autoresearch enabling AI research execution on Windows with RTX GPUs.
3/9/2026
Show HN: I built a BYOK AI gateway after a 100x Cloudflare KV cost mistake
Open source AI gateway built after cost issues, using Claude for automation in RSS-to-WordPress content pipeline.
3/9/2026
Show HN: Claude Toad scans your repo then generates your full Claude Code config
Tool that scans repositories and auto-generates Claude Code configuration files.
3/9/2026
Show HN: Bring your own prompts to remote shells
Tool that makes locally-defined prompts appear as executable commands in remote shells without SSH access for LLMs.
3/9/2026
If you are selling an AI Software product, read this
Analysis of competitive dynamics for AI product sellers in markets where LLMs are widely available.
3/9/2026
Show HN: Hebbs – a single-binary memory engine for AI agents – sub-10ms
Sub-10ms memory engine for AI agents implemented as single binary. Open source developer tool.
3/9/2026
Show HN: Sento AI – Generate landing pages from a text prompt (sento24.com)
Tool for generating landing pages from text prompts using AI.
3/9/2026
Show HN: ClawDiary – Audit log and human approval gate for AI agents (Claw, MCP
Audit log and approval gate for AI agents. Logs agent actions and blocks risky operations until user approval via Telegram.
3/9/2026
Show HN: Mallo – An iOS AI companion that remembers you and checks in daily
iOS app using LLM to extract and store long-term memories from conversations for personalized daily check-ins.
3/9/2026
Why LLM agents break when you give them 100k tools
Analysis of why LLM agents fail with large tool sets (100k+ tools) and practical limitations.
3/9/2026
Satsgate: A Lightning L402 paywall and prepaid credits for APIs/AI agents
Payment system using Bitcoin Lightning L402 protocol for API and AI agent access with prepaid credits.
3/9/2026
DropMind – a simple self-hosted memory cache
Self-hosted lightweight memory cache for personal notes, links, and files with privacy-first design for local deployment.
3/9/2026
Show HN: Crit – terminal TUI for reviewing AI-generated code and plan documents
Terminal TUI for inline review of AI-generated code and plans with comment feedback loop to agents.
3/9/2026
Show HN: Peribus – "Track hand on A, display on B, blink LED on C" – one prompt
Plan 9-inspired workspace where single prompts generate live UI and orchestrate hardware across networked machines via LLM.
3/9/2026
Building AI Coding Agents for the Terminal
Article title on AI coding agents for terminal use. Content missing; cannot assess quality.
3/9/2026
Show HN: Real-time violent action detection using computer vision
Real-time violence detection system combining object detection, pose estimation, and LSTM for suspicious action identification in video.
3/9/2026
Show HN: ClawAid – AI doctor that fixes OpenClaw in one command
Tool using Claude Sonnet to diagnose and auto-fix OpenClaw bugs by analyzing logs and system state. Runs locally with user confirmation.